As highly leveraged households are more sensitive to changes in income and interest rates,22 systematically significant shocks to the economy have a stronger impact on households’
financial stability and consumption in countries with high household debt. It could therefore be expected that economic crises have greater impact on countries with a higher potential drop in demand and investment as well as banking sector destabilisation due to higher loan delinquency.
Although involvement of household indebtedness in bringing about the double-dip recession that hit most of Europe in 2009, and subsequently in late 2011, was negligible, lower disposable income had a disproportionate impact on household consumption, since the ratio of debt service to income increased. In the same vein, the adjustment to new expected income reduced the usual consumption pattern and led to substantially lower domestic demand for goods and services, creating deflationary pressure and deepening the recession. In this theoretical framework, household debt overhang indeed has a clearly negative effect during crises and the largely positive economic effect of household debt expansion before the crisis has thus largely detrimental effects on consumption during the
22 Debelle, G. (2004), “Household debt and the macroeconomy”, BIS Quarterly Review, March, 51–64.
recession.23 In a context of fiscal consolidation and government unwillingness or incapacity to take on more debt to sustain the economic activity and smoothen the downturn, households’ deleveraging efforts even coincided with an economically difficult period during which investments were postponed as neither private nor public demand could guarantee returns, largely contributing to the fiscal-retrenchment multiplier recently adjusted by the IMF.24
While credit has a clearly positive effect on the short-term growth with positive spillovers for the future, generating substantial multipliers, it is also associated with long-term concerns of excessive debt accumulation. Even in economically peaceful times, excessive debt overhangs can strain on economy if the debt service becomes too detrimental to consumption. Research on the effects of a too high debt-to-GDP ratio has found a largely positive association between household credit and economic growth. Cecchetti et al. (2011)25 calculated a very approximate estimate of debt values upon which the debt can become detrimental to GDP growth. They observed a reversal in the positive association between household debt and GDP growth at a debt-to-GDP ratio of 85%. Such research has, however, recently been put in doubt26 and should therefore be considered with the utmost precaution. More importantly, the association between high debt to income and slower economic growth does not imply a direct causality.
In the European context, only Cyprus, Denmark, the Netherlands and the UK attain levels identified by Cecchetti et al., with Portugal and Spain ranking close below. More than nominal values of such limits per se, however, the main conclusion is that there are theoretical optimum debt levels – although most probably varying and relative – that are most likely to sustain maximum economic growth and that at a certain level, debt overhang could be detrimental to the economy. The most intuitive question would therefore be whether an immediate deleveraging would be the right thing to do in order to foster higher growth.
Such a conclusion would be rather doubtful, on the one hand due to the self-enforcing multiplier effects that household deleveraging has on demand and, on the other, due to the incapacity of the state to sustain demand through higher expenditure, as is often the case during less severe recessions. More importantly, as is mentioned in the subsequent chapters of this report, perceived sustainable levels of household debt can vary substantially between countries. It is therefore not clear where the debt overhang is of detrimental value to economic growth.
Furthermore, quick household-credit retrenchment under the recession can actually lead to an even deeper recession, thereby further destabilising household income and the extent of negative effects of household-debt overhang.27 Successful deleveraging after the build-up of debt overhangs generally proceeds during recovery (Roxburgh et al. 2012), which allows the household to partly reduce their nominal debt and partly grow out of the excessive leverage.
23 Dynan, K., A. Mian & K. Pence (2012), “Is A Household Debt Overhang Holding Back Consumption?”, Brookings Papers on Economic Activity.
24 Chmelar, A. (2012), “Household Debt in Europe’s Periphery: The dangers of a prolonged recession”, ECRI Commentary No. 12, 22 November.
25 Cecchetti, S., M. Mohanty & F. Zampolli (2011), “The real effects of debt”, pp. 25–27.
26 Herndon, T., M. Ash & R. Pollin (2013), “Does high public debt consistently stifle economic growth? A critique of Reinhart and Rogoff”, Political Economy Research Institute Working Paper Series, 322.
27 Chmelar, A. (2012), “Household Debt in Europe’s Periphery: The dangers of a prolonged recession”, ECRI Commentary No. 12, 22 November.
A different way for the household sector to reduce their debt is to shift the cost onto the financial sector and ultimately onto the government or the lender of last resort. In countries with a low enforceability of evictions and more debtor-centred personal insolvency rules, a non-negligible part of the debt overhang was eliminated through personal defaults. Unlike Spain and Portugal, where the ratio of delinquent loans was kept at very low levels between 2 and 4%, in Ireland and Greece this de facto elimination occurred due to high non-performing loan ratios of 12 and 14% respectively.28 While the effect on consumption in Ireland was potentially tangible and could have helped the household sector to pick up, the context of the deep recession in Greece does not allow a similar conclusion.
4.1 Perceived household-debt sustainability levels
It is a stylised fact that there is a relationship between debt sustainability and the output of the economy or the disposable income of households. As expectations of future growth decrease, we might anticipate a lowering of debt to income, especially in countries with high debt-to-income ratios. This is, however, not what is happening during the ongoing deleveraging episode.
In the past four years of low growth and recessions, households with the biggest leverage in the EU27 do not necessarily reduce their debt levels more than those with a lower one. The relationship between debt reduction as a reaction to the crisis and the overall leverage levels in the EU is statistically insignificant (see Figure 10). This implies a counterintuitive conclusion that, at least in macroeconomic terms, households do not reduce their debt because they are too indebted. Instead, the main indicator of debt-overhang reduction appears to be its recent build-up with a highly significant correlation.
Figure 10. Rates of household-debt growth during the crisis (2009-2012) in relation to debt pre-crisis expansion (2003-2007) and to household debt-to-GDI ratio
This sheds light on two major problems of household debt development before and during the crisis. It is not the level of debt that determines the necessity to reduce the debt; it is rather the perception of the sustainability of debt, determined significantly by the time when the debt has built up. It also shows that debt-reduction cannot be dictated or determined by
Leverage to GDI at peak (December 2007) Linear prediction plot (R²= .0024) Source: ECRI, ECB, Eurostat; Note: *** p<0.01
pointing to single ratios to GDP or to disposable income. The perceived or real sustainability of debt levels depend on different variables than income per se.
Two conclusions could be drawn from this argumentation. Either the debt levels were sustainable but the culture of higher debt did not have time to adjust to new income expectations, or the exuberance of credit expansion before the crisis was irrational and the credit taken out was not sufficiently consolidated in household portfolios. It does not imply that the bulk of the credit expansion before the crisis was speculative in nature, but rather that it was the result of an over-optimistic reaction to lower credit constraints and that indeed could be depicted as subconsciously speculating on – ex post unrealistic – future income.